CN109753880B - Detection and identification method for natural scene vehicle-mounted video road speed limit sign - Google Patents
Detection and identification method for natural scene vehicle-mounted video road speed limit sign Download PDFInfo
- Publication number
- CN109753880B CN109753880B CN201811521604.5A CN201811521604A CN109753880B CN 109753880 B CN109753880 B CN 109753880B CN 201811521604 A CN201811521604 A CN 201811521604A CN 109753880 B CN109753880 B CN 109753880B
- Authority
- CN
- China
- Prior art keywords
- speed limit
- limit sign
- road speed
- image
- vehicle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a detection and identification method of a vehicle-mounted video road speed limit sign in a natural scene, which comprises the following steps: firstly, collecting an image containing a road speed limit sign from a vehicle-mounted video of a natural scene; then, sequentially carrying out filtering processing, threshold segmentation, binarization processing and morphological processing on the image containing the road speed limit sign; then, positioning the road speed limit sign in the image after image processing to obtain the area of the road speed limit sign in the image; then, character segmentation is carried out on the area of the road speed limit sign in the image to obtain all characters in the road speed limit sign; and finally, recognizing all the segmented characters, namely finishing the recognition of the road speed limit sign. The speed limit identification method and the speed limit identification device can be suitable for video recording equipment such as the existing automobile data recorder, have wide application range, do not need to add new hardware facilities, have low use cost, can well identify the speed limit identification aiming at images collected under various complicated road conditions or extreme weather, and have high identification accuracy.
Description
Technical Field
The invention belongs to the field of target detection and identification, and particularly relates to a method for detecting and identifying a vehicle-mounted video road speed limit sign in a natural scene.
Background
The detection and identification of the natural scene vehicle-mounted video road speed limit signs mainly aim at the identification of a plurality of speed limit signs in a vehicle-mounted camera video in the driving process, images are obtained through the vehicle-mounted camera, and the road speed limit signs are identified based on an image processing technology, so that the road speed limit signs are quickly and accurately identified, and accurate information can be timely provided for traffic navigation. The method has important significance for the research of the detection and identification method of the road speed limit sign, is widely concerned by researchers at home and abroad, has great practical significance and practical value, can reduce the occurrence of traffic accidents, and improves the harmonious stability of the society.
After a long time of development, the image processing technology has a relatively perfect theory and is widely used in various fields, but at present, few papers and patents are available in the field of image recognition of road speed limit signs. The camera pictures of the road speed limit signs have the characteristics of distortion, low pixels and the like, so that the identification difficulty is increased, and when the road speed limit signs face various extreme weathers, the accuracy of detection and identification of the speed limit signs is low, the accuracy of identification of specific speed limit values is not high, and the like, so that the requirements of detection and identification of the road speed limit signs are difficult to meet. To solve the problem, duyili and the like propose a method for converting an original image into a brightness space, and then positioning a speed limit sign by adopting a shape edge detection algorithm [ duyili, jiayonghong. 32-34], the method has certain disadvantages that the adopted edge detection algorithm is easily interfered by arc noise, thereby causing false detection, and the detection success rate is not high.
Disclosure of Invention
The invention aims to provide a method for detecting and identifying a road speed limit sign in a vehicle-mounted video in a natural scene.
The technical solution for realizing the purpose of the invention is as follows: a detection and identification method for a natural scene vehicle-mounted video road speed limit sign comprises the following steps:
step 1, collecting images containing road speed limit signs from a vehicle-mounted video of a natural scene;
step 2, filtering the image containing the road speed limit sign;
step 3, performing threshold segmentation on the image subjected to filtering processing in the step 2;
step 4, carrying out binarization processing on the image subjected to threshold segmentation in the step 3;
step 5, performing morphological processing and filtering processing on the binary image obtained in the step 4;
step 6, positioning the road speed limit sign in the image processed in the step 5 to obtain the area of the road speed limit sign in the image;
step 7, carrying out character segmentation on the area of the road speed limit sign in the image obtained in the step 6 to obtain all characters in the road speed limit sign;
and 8, recognizing all the characters segmented in the step 7, namely finishing the recognition of the road speed limit sign.
Compared with the prior art, the invention has the following remarkable advantages: 1) the method can be suitable for the existing video equipment such as the automobile data recorder and the like, has wide application range, does not need to add new hardware facilities, and has low use cost; 2) according to the invention, an image segmentation method based on an HSV color space is adopted, and image segmentation can be well realized aiming at images collected under various complex road conditions or extreme weather; 3) according to the invention, the image is processed through the filtering algorithm, the Hough transformation algorithm, the image filling algorithm and other algorithms, the positioning precision is higher, and the identification efficiency is effectively improved; 4) the invention adopts a block-based feature matching method, thereby improving the accuracy of character recognition.
The present invention is described in further detail below with reference to the attached drawing figures.
Drawings
FIG. 1 is a flow chart of the detection and identification method of the vehicle-mounted video road speed limit sign in the natural scene.
Fig. 2 is an image including a road speed limit sign captured from a natural scene vehicle-mounted video according to an embodiment of the present invention.
FIG. 3 is an image after thresholding in an embodiment of the invention.
Fig. 4 is an image after binarization processing is performed on the image after threshold segmentation in the embodiment of the present invention.
Fig. 5 is an image after the morphological processing and the filtering processing are performed on the binarized image according to the embodiment of the present invention.
Fig. 6 is an image of a circular target area in an image after extraction processing according to an embodiment of the present invention.
Fig. 7 is an image of all the characters in the road speed limit sign obtained in the embodiment of the present invention.
Detailed Description
With reference to fig. 1, the invention provides a method for detecting and identifying a natural scene vehicle-mounted video road speed limit sign, which comprises the following steps:
step 1, collecting images containing road speed limit signs from a vehicle-mounted video of a natural scene.
And 2, filtering the image containing the road speed limit sign to eliminate obvious noise.
Further, the filtering process specifically adopts a median filtering process.
And 3, performing threshold segmentation on the image subjected to the filtering processing in the step 2.
Further, the threshold segmentation specifically adopts an image segmentation method based on an HSV color space.
And 4, carrying out binarization processing on the image subjected to threshold segmentation in the step 3.
And 5, performing morphological processing and filtering processing on the binary image obtained in the step 4 to eliminate noise and enhance the connectivity of a connected region.
And 6, positioning the road speed limit sign in the image processed in the step 5 to obtain the area of the road speed limit sign in the image.
Further, the road speed limit sign in the image processed in the step 5 is located to obtain an area of the road speed limit sign in the image, and the method specifically comprises the following steps:
step 6-1, extracting all circular target areas in the image processed in the step 5 by using a Hough circle transformation detection method;
6-2, vertically projecting pixels in each circular target area by using a projection method to obtain a plurality of vertical projections, and horizontally projecting the target area corresponding to each vertical projection;
6-3, extracting the circular target area corresponding to the road speed limit sign according to the number of vertical projections after the vertical projection of the pixels in each circular target area, the relation between the width of each vertical projection and the diameter of the circular target area where the vertical projection is located, and the height of the projection after the horizontal projection of the target area corresponding to each vertical projection, wherein the specific steps are as follows:
if the pixels in the circular target area are subjected to vertical projection and horizontal projection in sequence, the following conditions are met simultaneously: the number of the vertical projections is 1-3, the width of each vertical projection is 1/4-1/3 of the diameter of the circular target area where the vertical projection is located, and the heights of all the horizontal projections are 1/3-1/2 of the diameter of the circular target area where the horizontal projection is located, so that the circular target area is the road speed limit sign.
Further, histogram equalization and geometric normalization processing are carried out on the image processed in the step 5 before the step 6 is executed, so that the problems of color distortion and geometric deformation of the detected speed limit sign can be solved.
And 7, performing character segmentation on the area of the road speed limit sign in the image obtained in the step 6 to obtain all characters in the road speed limit sign.
And 8, recognizing all the characters segmented in the step 7, namely finishing the recognition of the road speed limit sign.
Further, all the characters segmented in the step 7 are recognized and combined, that is, the recognition of the road speed limit sign is completed, which specifically comprises the following steps:
step 8-1, carrying out normalization processing on each character segmented in the step 7 to make the character size consistent with that of the character in the template library;
and 8-2, respectively matching each character to be recognized after normalization processing in the step 8-1 with all characters in the template library by using an image matching algorithm in a sequence from left to right, and solving corresponding correlation coefficients, wherein the character in the template library corresponding to the largest correlation coefficient is the character to be recognized, so that the recognition of the road speed limit sign is completed.
Further, the image matching algorithm in the step 8-2 specifically adopts a block-based feature matching method.
Furthermore, the block-based feature matching method is a block-based SIFT feature matching method.
Further, the block-based feature matching method is a block-based SURF feature matching method.
Examples
The invention relates to a detection and identification method of a natural scene vehicle-mounted video road speed limit sign, which comprises the following contents:
1. an image containing a road speed limit sign is collected from a natural scene vehicle-mounted video as shown in fig. 2.
2. And carrying out median filtering processing on the acquired image to remove obvious noise.
3. And performing threshold segmentation on the filtered image by adopting an image segmentation method based on an HSV color space, as shown in FIG. 3.
4. The binarized image after the threshold segmentation is subjected to binarization processing as shown in fig. 4.
5. The binarized image is subjected to morphological processing and filtering processing as shown in fig. 5.
6. And (4) positioning the road speed limit sign in the image after the processing of the step 5 by utilizing a Hough circle conversion detection method to obtain the area of the road speed limit sign in the image, wherein the area is shown in figure 6.
7. And carrying out character segmentation on the area of the road speed limit sign in the image to obtain all characters in the road speed limit sign.
8. Recognizing all the segmented characters, namely recognizing the road speed limit sign, specifically, normalizing each segmented character to make the size of the character consistent with that of the characters in the template library, then respectively matching each character to be recognized after normalization processing with all the characters in the template library by using an image matching algorithm in a sequence from left to right based on a block SURF feature matching method, and solving a corresponding correlation coefficient, wherein the character in the template library corresponding to the largest correlation coefficient is the character to be recognized, thereby completing the recognition of the road speed limit sign, as shown in FIG. 7.
The identification experiment is carried out on 103 frames of video images of 112 speed limit signs, HSV and characteristic vector character blocking template matching, Hough and projection methods and other related algorithms and improved algorithms are adopted, the identification accuracy rate reaches 97.56%, and the traditional method is 85.41%. The result shows that the method greatly improves the identification accuracy of the road speed limit sign and has high application value.
The invention can be suitable for the existing video equipment such as the automobile data recorder and the like, has wide application range, does not need to add new hardware facilities and has low use cost. The speed limit sign recognition method can well recognize the speed limit sign aiming at the images acquired under various complex road conditions or extreme weather, and has high recognition accuracy.
Claims (8)
1. A detection and identification method for a vehicle-mounted video road speed limit sign in a natural scene is characterized by comprising the following steps:
step 1, collecting an image containing a road speed limit sign from a vehicle-mounted video of a natural scene;
step 2, filtering the image containing the road speed limit sign;
step 3, carrying out threshold segmentation on the image subjected to filtering processing in the step 2;
step 4, carrying out binarization processing on the image subjected to threshold segmentation in the step 3;
step 5, performing morphological processing and filtering processing on the binary image obtained in the step 4;
step 6, positioning the road speed limit sign in the image processed in the step 5 to obtain the area of the road speed limit sign in the image; the method specifically comprises the following steps:
step 6-1, extracting all circular target areas in the image processed in the step 5 by using a Hough circle transformation detection method;
6-2, vertically projecting pixels in each circular target area by using a projection method to obtain a plurality of vertical projections, and horizontally projecting the target area corresponding to each vertical projection;
6-3, extracting the circular target area corresponding to the road speed limit sign according to the number of vertical projections after the vertical projection of the pixels in each circular target area, the relation between the width of each vertical projection and the diameter of the circular target area where the vertical projection is located, and the height of the projection after the horizontal projection of the target area corresponding to each vertical projection, wherein the specific steps are as follows:
if the pixels in the circular target area are subjected to vertical projection and horizontal projection in sequence, the following conditions are met simultaneously: the number of the vertical projections is 1-3, the width of each vertical projection is 1/4-1/3 of the diameter of the circular target area where the vertical projection is located, and the heights of all the horizontal projections are 1/3-1/2 of the diameter of the circular target area where the horizontal projection is located, so that the circular target area is the road speed limit sign;
step 7, carrying out character segmentation on the area of the road speed limit sign in the image obtained in the step 6 to obtain all characters in the road speed limit sign;
and 8, recognizing all the characters segmented in the step 7, namely finishing the recognition of the road speed limit sign.
2. The method for detecting and identifying the speed limit sign of the vehicle-mounted video road in the natural scene according to claim 1, wherein the filtering processing in the step 2 specifically adopts median filtering processing.
3. The method for detecting and identifying the vehicle-mounted video road speed limit sign in the natural scene according to claim 1, wherein the threshold segmentation in the step 3 is carried out by adopting an image segmentation method based on an HSV color space.
4. The method for detecting and identifying the vehicle-mounted video road speed limit sign in the natural scene according to claim 1, wherein histogram equalization and geometric normalization processing are performed on the image processed in the step 5 before step 6 is performed.
5. The method for detecting and identifying the natural scene vehicle-mounted video road speed limit sign according to claim 1, wherein the step 8 of identifying and combining all the characters segmented in the step 7 is to complete identification of the road speed limit sign, and specifically comprises the following steps:
step 8-1, carrying out normalization processing on each character segmented in the step 7 to enable the size of each character to be consistent with that of the character in the template library;
and 8-2, respectively matching each character to be recognized after normalization processing in the step 8-1 with all characters in the template library by using an image matching algorithm in a sequence from left to right, and solving corresponding correlation coefficients, wherein the character in the template library corresponding to the largest correlation coefficient is the character to be recognized, so that the recognition of the road speed limit sign is completed.
6. The method for detecting and identifying the speed limit sign of the vehicle-mounted video road in the natural scene according to claim 5, wherein the image matching algorithm in the step 8-2 specifically adopts a block-based feature matching method.
7. The method for detecting and identifying the speed-limiting sign of the vehicle-mounted video road with the natural scene as claimed in claim 6, wherein the block-based feature matching method in the step 8-2 is a block-based SIFT feature matching method.
8. The method for detecting and identifying the speed limit sign of the vehicle-mounted video road in the natural scene as claimed in claim 6, wherein the block-based feature matching method in the step 8-2 is a block-based SURF feature matching method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811521604.5A CN109753880B (en) | 2018-12-13 | 2018-12-13 | Detection and identification method for natural scene vehicle-mounted video road speed limit sign |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811521604.5A CN109753880B (en) | 2018-12-13 | 2018-12-13 | Detection and identification method for natural scene vehicle-mounted video road speed limit sign |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109753880A CN109753880A (en) | 2019-05-14 |
| CN109753880B true CN109753880B (en) | 2022-09-27 |
Family
ID=66402613
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811521604.5A Active CN109753880B (en) | 2018-12-13 | 2018-12-13 | Detection and identification method for natural scene vehicle-mounted video road speed limit sign |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109753880B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110334601A (en) * | 2019-06-04 | 2019-10-15 | 武汉极目智能技术有限公司 | A kind of speed(-)limit sign board recognition methods of combination machine learning and computer vision |
| CN113792611A (en) * | 2021-08-20 | 2021-12-14 | 昆明理工大学 | Method for identifying road speed limit sign running on auxiliary driving edge equipment |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102968624A (en) * | 2012-12-12 | 2013-03-13 | 天津工业大学 | Method for positioning human eyes in human face image |
| CN104834932A (en) * | 2015-04-29 | 2015-08-12 | 河南城建学院 | Matlab algorithm of automobile license plate identification |
-
2018
- 2018-12-13 CN CN201811521604.5A patent/CN109753880B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102968624A (en) * | 2012-12-12 | 2013-03-13 | 天津工业大学 | Method for positioning human eyes in human face image |
| CN104834932A (en) * | 2015-04-29 | 2015-08-12 | 河南城建学院 | Matlab algorithm of automobile license plate identification |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109753880A (en) | 2019-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020173022A1 (en) | Vehicle violation identifying method, server and storage medium | |
| WO2018018788A1 (en) | Image recognition-based meter reading apparatus and method thereof | |
| Shidore et al. | Number plate recognition for indian vehicles | |
| CN104361336A (en) | Character recognition method for underwater video images | |
| KR20130105952A (en) | Method and apparatus for vehicle license plate recognition | |
| Paunwala et al. | A novel multiple license plate extraction technique for complex background in Indian traffic conditions | |
| CN102799859A (en) | Method for identifying traffic sign | |
| CN111027544B (en) | MSER license plate positioning method and system based on visual saliency detection | |
| CN103324958B (en) | Based on the license plate locating method of sciagraphy and SVM under a kind of complex background | |
| CN105718916A (en) | Lane line detection method based on Hough transform | |
| CN109753880B (en) | Detection and identification method for natural scene vehicle-mounted video road speed limit sign | |
| CN111915583A (en) | Vehicle and pedestrian detection method based on vehicle-mounted thermal infrared imager in complex scene | |
| CN104616019A (en) | Identification method for electronic equipment signboard image | |
| CN108520252B (en) | Road sign identification method based on generalized Hough transform and wavelet transform | |
| CN114998815A (en) | Traffic vehicle identification tracking method and system based on video analysis | |
| Vinoharan et al. | A wheel-based side-view car detection using snake algorithm | |
| CN103065137A (en) | License plate character recognition method | |
| Huang | A novel video text extraction approach based on Log-Gabor filters | |
| Liu et al. | Real-time speed limit sign detection and recognition from image sequences | |
| Sathya et al. | Vehicle license plate recognition (vlpr) | |
| CN204347862U (en) | A kind of vehicle vehicle-logo recognition system | |
| Dandu et al. | Vehicular number plate recognition using edge detection and characteristic analysis of national number plates | |
| Tian et al. | Robust traffic sign detection in complex road environments | |
| CN113449629A (en) | Lane line false and true identification device, method, equipment and medium based on driving video | |
| CN112418143A (en) | Traffic sign identification method for unmanned vehicle |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |